基于BP神经网络PID控制的风机风量控制研究  

Research on fan air volume control based on BP neural network PID control

作　　者：周爽 张春雨 陈伟 ZHOU Shuang;ZHANG Chunyu;CHEN Wei(College of Mechanical Engineering,Anhui Science and Technology University,Fengyang 233100,China)

机构地区：[1]安徽科技学院机械工程学院,安徽凤阳233100

出　　处：《安徽科技学院学报》2023年第6期89-95,共7页Journal of Anhui Science and Technology University

基　　金：国家自然科学基金(32002123);安徽省科技重大专项(18030701198)。

摘　　要：目的:为提高好氧堆肥转化率,改善常规PID算法在好氧堆肥风机风量控制中的时滞性、非线性等问题。方法:提出BP神经网络PID算法控制策略,结合BP神经网络算法与传统PID算法,实现堆肥风机风量控制参数在线自整定。在MATLAB软件Simulink环境下搭建常规PID、模糊PID与BP神经网络PID模型,对比分析3种控制算法对风机风量控制的稳定性,并在300 s时添加干扰信号验证系统稳定性。结果:在BP神经网络PID控制下,系统超调量减小,与另外2种控制方法相比稳态误差分别降低13%和8%,稳态时间缩短70 s。结论:BP神经网络PID控制算法系统响应速度快、调节时间短且稳定,鲁棒性好,抗干扰能力强,为好氧堆肥风机风量控制提供参考。Objective:To improve the conversion rate of aerobic composting,improve the problems of time delay and nonlinearity of conventional PID algorithm in the air volume control of aerobic compost fan.Methods:A control strategy based on BP neural network PID algorithm was proposed,and the BP neural network algorithm was combined with traditional PID algorithm to realize online self-tuning of compost fan air volume control parameters.In the MATLAB software Simulink environment,the conventional PID,fuzzy PID and BP neural network PID models were built,the stability of the three control algorithms on the air volume control of the fan was compared and analyzed,and the interference signal was added at 300 s to verify the stability of the system.Results:Under the control of BP neural network PID,the system overshoot was reduced,the error was reduced by 8%compared with the other two control methods,and the steady-state time is shortened by 70 s.Conclusion:The BP neural network PID control algorithm system had fast response speed,short and stable adjustment time,good robustness and strong anti-interference ability,which provided a reference for the air volume control of aerobic compost fan.

关 键 词：BP神经网络 PID控制 风机风量 好氧堆肥 

分 类 号：S815[农业科学—畜牧学] X713[农业科学—畜牧兽医]